Research On New Methods Of Parameters Estimation And Signal Simulation For SAR

Doppler parameters estimation and raw signal simulation for SAR (synthetic aperture radar) have attracted much attention in SAR technical field. The former is necessary for SAR high-quality image formation, while the latter for SAR technical study and system development. The methods for resolving SAR PRF-ambiguity, jointly estimating Doppler parameters for squinted SAR, blindly imaging SAR data independent of apriori knowledge about SAR motion, and fast simulation of SAR raw signal are researched in this thesis. The proposed algorithms about parameters estimation are verified with real SAR data, and the raw signal simulation method has been applied in a real SAR simulation system. The innovations of this thesis are as follows:1. The approximate linear relation between SAR Doppler centroid and the slant range of the scene is derived. Based on the relation, a novel method for resolving SAR PRF-ambiguity is proposed. Implemented in the two-dimensional temporal domain, the method is suit for low contrast scene, and computationally efficient with only the estimation of ambiguous Doppler centroid needed.2. The statistical property of SAR signal in the range Doppler (RD) domain is researched. By using the statistical property, two robust methods to resolve PRF-ambiguity: RD domain contrast maximization and minimization,are proposed.3. SAR squint mode would degrade the performance of traditional auto-focus methods. To solve this problem, a joint estimation approach of Doppler parameters, based on RD domain contrast maximization and minimization, is presented, which efficiently improves the auto-focus accuracy.4. On the basis of joint estimation method of Doppler parameters, and the improvement on traditional phase difference auto-focus method, SAR blind image formation algorithm is proposed. The algorithm can image SAR data,being free of any apriori information about SAR platform motion.5. Fast algorithm for SAR raw signal simulation is proposed. By taking wavenumber domain imaging algorithm as an example, the reversible relationship between SAR signal simulating algorithm and SAR imaging algorithm is derived. For the SAR that transmits LFM signal, it is proved that the forward method which simulates SAR receiving process is consistent with the backward method that utilizes the inverse process of SAR imaging algorithm. Fast implementation of Shannon interpolation is presented.6. By using DBS, a method to insert the SAR platform motion position deviation in the simulated raw signal is proposed, which simulates the space-variance of the effect of the position deviation on SAR raw signal.